Merge pull request Sage-Bionetworks#184 from itismeghasyam/vignette-d…

…evelop

Vignette develop

R/data.R

@@ -113,24 +113,20 @@
 #' Sample heartrate data from a smartphone camera
 #' 
 #' A dataframe containing sample JSON output format of the heartrate data,
-#' containing red, green, blue levels from the camera sensor, along with the
-#' timestamp, and the derived metrics of hue, saturation and brightness.
+#' containing red, green, blue levels from the camera sensor, indexed with the
+#' timestamp t.
 #' 
 #' A video(240p) of a finger covering the smartphone camera with the flash
 #' turned on was recorded at 60fps. The average intensity of all the pixels per
 #' frame in all the three color channels was calculated, and indexed along
-#' the timestamp of the frame. We also have hue, brightness and saturation
-#' calculated in a similar way.
+#' the timestamp of the frame.
 #' 
 #' @format A data frame with 3661 rows (observations) and 7 variables:
 #' \describe{
 #'   \item{t}{time, in seconds}
 #'   \item{green}{mean green intensity across the frame at the given timestamp}
 #'   \item{blue}{mean blue intensity across the frame at the given timestamp}
 #'   \item{red}{mean red intensity across the frame at the given timestamp}
-#'   \item{brightness}{mean brightness across the frame at the given timestamp}
-#'   \item{hue}{mean hue across the frame at the given timestamp}
-#'   \item{saturation}{mean saturation across the frame at the given timestamp}
 #' }
 "heartrate_data"
 

R/get_balance_features.R

@@ -67,37 +67,27 @@
 #' @examples 
 #' accelerometer_data = cbind(
 #'   t = balance_data$timestamp,
-#'   balance_data$userAcceleration)
+#'   balance_data$userAcceleration) %>% 
+#'   dplyr::filter(t < 5)
 #' gyroscope_data = cbind(
 #'   t = balance_data$timestamp,
-#'   balance_data$rotationRate)
+#'   balance_data$rotationRate) %>% 
+#'   dplyr::filter(t < 5)
 #' 
 #' balance_features <- get_balance_features(accelerometer_data, gyroscope_data)
 #' 
 #' balance_features <- get_balance_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
-#'   time_filter = c(2,8))
-#' 
-#' balance_features <- get_balance_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   detrend = TRUE)
-#' 
-#' balance_features <- get_balance_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   frequency_filter = c(0.5, 25))
-#' 
-#' balance_features <- get_balance_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   window_length = 512,
-#'   window_overlap = 0.9)
+#'   time_filter = c(2,4),
+#'   detrend = TRUE,
+#'   frequency_filter = c(0.5,25))
 #' 
 #' balance_features <- get_balance_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
+#'   window_length = 256,
+#'   window_overlap = 0.2,
 #'   derived_kinematics = TRUE)
 #' 
 #' balance_features <- get_balance_features(

R/get_heartrate.R

@@ -43,6 +43,9 @@ get_heartrate <- function(heartrate_data, window_length = 10, window_overlap = 0
   # Convert window length from seconds to samples
   window_length <- round(sampling_rate * window_length)
   mean_filter_order <- 65
+  if(sampling_rate <= 32){
+    mean_filter_order <- 33
+    }
 
   ##  Apply pre-processing filter to all heartrate data
 
@@ -222,10 +225,11 @@ get_hr_from_time_series <- function(x, sampling_rate, method = 'acf', min_hr = 4
     confidence <- NA
   }
 
+
   # If hr or condidence is NaN, then return hr = 0 and confidence = 0
-  if (is.na(confidence) || is.na(hr)) {
-    confidence <- NA
-    hr <- NA
+  if ((length(hr) == 0) || is.null(hr)) {
+  confidence <- NA
+  hr <- NA
   }
 
   return(c(hr, confidence))

R/get_kinetic_tremor_features.R

@@ -76,10 +76,12 @@
 #' @examples 
 #' accelerometer_data = cbind(
 #'   t = kinetic_tremor_data$timestamp,
-#'   kinetic_tremor_data$userAcceleration)
+#'   kinetic_tremor_data$userAcceleration) %>% 
+#'   dplyr::filter(t < 5)
 #' gyroscope_data = cbind(
 #'   t = kinetic_tremor_data$timestamp,
-#'   kinetic_tremor_data$rotationRate)
+#'   kinetic_tremor_data$rotationRate) %>% 
+#'   dplyr::filter(t < 5)
 #' 
 #' kinetic_tremor_features <- get_kinetic_tremor_features(
 #'   accelerometer_data,
@@ -88,27 +90,15 @@
 #' kinetic_tremor_features <- get_kinetic_tremor_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
-#'   time_filter = c(2,8))
-#' 
-#' kinetic_tremor_features <- get_kinetic_tremor_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   detrend = TRUE)
-#' 
-#' kinetic_tremor_features <- get_kinetic_tremor_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
+#'   time_filter = c(2,4),
+#'   detrend = TRUE,
 #'   frequency_filter = c(0.5, 25))
 #' 
 #' kinetic_tremor_features <- get_kinetic_tremor_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
-#'   window_length = 512,
-#'   window_overlap = 0.9)
-#' 
-#' kinetic_tremor_features <- get_kinetic_tremor_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
+#'   window_length = 256,
+#'   window_overlap = 0.2,
 #'   derived_kinematics = TRUE)
 #' 
 #' kinetic_tremor_features <- get_kinetic_tremor_features(

R/get_tremor_features.R

@@ -75,37 +75,27 @@
 #' @examples 
 #' accelerometer_data = cbind(
 #'   t = rest_tremor_data$timestamp,
-#'   rest_tremor_data$userAcceleration)
+#'   rest_tremor_data$userAcceleration) %>% 
+#'   dplyr::filter(t < 5)
 #' gyroscope_data = cbind(
 #'   t = rest_tremor_data$timestamp,
-#'   rest_tremor_data$rotationRate)
+#'   rest_tremor_data$rotationRate) %>% 
+#'   dplyr::filter(t < 5)
 #' 
 #' tremor_features <- get_tremor_features(accelerometer_data, gyroscope_data)
 #' 
 #' tremor_features <- get_tremor_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
-#'   time_filter = c(2,8))
-#' 
-#' tremor_features <- get_tremor_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   detrend = TRUE)
-#' 
-#' tremor_features <- get_tremor_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
+#'   time_filter = c(2,4),
+#'   detrend = TRUE,
 #'   frequency_filter = c(0.5, 25))
 #' 
 #' tremor_features <- get_tremor_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
-#'   window_length = 512,
-#'   window_overlap = 0.9)
-#' 
-#' tremor_features <- get_tremor_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
+#'   window_length = 256,
+#'   window_overlap = 0.2,
 #'   derived_kinematics = TRUE)
 #' 
 #' tremor_features <- get_tremor_features(

R/get_walk_features.R

@@ -68,10 +68,12 @@
 #' @examples 
 #' accelerometer_data = cbind(
 #'   t = walk_data$timestamp,
-#'   walk_data$userAcceleration)
+#'   walk_data$userAcceleration) %>% 
+#'   dplyr::filter(t < 5)
 #' gyroscope_data = cbind(
 #'   t = walk_data$timestamp,
-#'   walk_data$rotationRate)
+#'   walk_data$rotationRate) %>% 
+#'   dplyr::filter(t < 5)
 #' 
 #' walk_features <- get_walk_features(
 #'   accelerometer_data,
@@ -80,27 +82,15 @@
 #' walk_features <- get_walk_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
-#'   time_filter = c(2,8))
-#' 
-#' walk_features <- get_walk_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   detrend = TRUE)
-#' 
-#' walk_features <- get_walk_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   frequency_filter = c(0.5, 25))
-#' 
-#' walk_features <- get_walk_features(
-#'   accelerometer_data,
-#'   gyroscope_data,
-#'   window_length = 512,
-#'   window_overlap = 0.9)
+#'   time_filter = c(2,4),
+#'   detrend = TRUE,
+#'   frequency_filter = c(0.5,25))
 #' 
 #' walk_features <- get_walk_features(
 #'   accelerometer_data,
 #'   gyroscope_data,
+#'   window_length = 256,
+#'   window_overlap = 0.2,
 #'   derived_kinematics = TRUE)
 #' 
 #' walk_features <- get_walk_features(